US20040255211A1 - Apparatus and method for reducing test resources in testing DRAMs - Google Patents

Apparatus and method for reducing test resources in testing DRAMs Download PDF

Info

Publication number
US20040255211A1
US20040255211A1 US10/853,573 US85357304A US2004255211A1 US 20040255211 A1 US20040255211 A1 US 20040255211A1 US 85357304 A US85357304 A US 85357304A US 2004255211 A1 US2004255211 A1 US 2004255211A1
Authority
US
United States
Prior art keywords
control signal
bank
memory device
address
locations
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
US10/853,573
Other versions
US7168018B2 (en
Inventor
Chris Cooper
Siang Giam
Jerry McBride
Scott Gatzemeier
Scott Ayres
David Brown
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Individual
Original Assignee
Individual
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Individual filed Critical Individual
Priority to US10/853,573 priority Critical patent/US7168018B2/en
Publication of US20040255211A1 publication Critical patent/US20040255211A1/en
Priority to US11/445,944 priority patent/US20070168790A1/en
Application granted granted Critical
Publication of US7168018B2 publication Critical patent/US7168018B2/en
Adjusted expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C29/00Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
    • G11C29/006Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation at wafer scale level, i.e. wafer scale integration [WSI]
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/28Testing of electronic circuits, e.g. by signal tracer
    • G01R31/317Testing of digital circuits
    • G01R31/3181Functional testing
    • G01R31/319Tester hardware, i.e. output processing circuits
    • G01R31/3193Tester hardware, i.e. output processing circuits with comparison between actual response and known fault free response
    • G01R31/31932Comparators
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C29/00Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
    • G11C29/04Detection or location of defective memory elements, e.g. cell constructio details, timing of test signals
    • G11C29/08Functional testing, e.g. testing during refresh, power-on self testing [POST] or distributed testing
    • G11C29/12Built-in arrangements for testing, e.g. built-in self testing [BIST] or interconnection details
    • G11C29/1201Built-in arrangements for testing, e.g. built-in self testing [BIST] or interconnection details comprising I/O circuitry
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C29/00Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
    • G11C29/04Detection or location of defective memory elements, e.g. cell constructio details, timing of test signals
    • G11C29/08Functional testing, e.g. testing during refresh, power-on self testing [POST] or distributed testing
    • G11C29/12Built-in arrangements for testing, e.g. built-in self testing [BIST] or interconnection details
    • G11C29/12015Built-in arrangements for testing, e.g. built-in self testing [BIST] or interconnection details comprising clock generation or timing circuitry
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C29/00Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
    • G11C29/04Detection or location of defective memory elements, e.g. cell constructio details, timing of test signals
    • G11C29/08Functional testing, e.g. testing during refresh, power-on self testing [POST] or distributed testing
    • G11C29/12Built-in arrangements for testing, e.g. built-in self testing [BIST] or interconnection details
    • G11C29/14Implementation of control logic, e.g. test mode decoders
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C29/00Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
    • G11C29/04Detection or location of defective memory elements, e.g. cell constructio details, timing of test signals
    • G11C29/08Functional testing, e.g. testing during refresh, power-on self testing [POST] or distributed testing
    • G11C29/12Built-in arrangements for testing, e.g. built-in self testing [BIST] or interconnection details
    • G11C29/18Address generation devices; Devices for accessing memories, e.g. details of addressing circuits
    • G11C29/26Accessing multiple arrays
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C29/00Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
    • G11C29/04Detection or location of defective memory elements, e.g. cell constructio details, timing of test signals
    • G11C29/08Functional testing, e.g. testing during refresh, power-on self testing [POST] or distributed testing
    • G11C29/12Built-in arrangements for testing, e.g. built-in self testing [BIST] or interconnection details
    • G11C29/38Response verification devices
    • G11C29/40Response verification devices using compression techniques
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C29/00Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
    • G11C29/04Detection or location of defective memory elements, e.g. cell constructio details, timing of test signals
    • G11C29/08Functional testing, e.g. testing during refresh, power-on self testing [POST] or distributed testing
    • G11C29/48Arrangements in static stores specially adapted for testing by means external to the store, e.g. using direct memory access [DMA] or using auxiliary access paths
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C29/00Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
    • G11C29/56External testing equipment for static stores, e.g. automatic test equipment [ATE]; Interfaces therefor
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C11/00Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor
    • G11C11/21Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements
    • G11C11/34Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices
    • G11C11/40Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices using transistors
    • G11C11/401Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices using transistors forming cells needing refreshing or charge regeneration, i.e. dynamic cells
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C29/00Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
    • G11C29/56External testing equipment for static stores, e.g. automatic test equipment [ATE]; Interfaces therefor
    • G11C2029/5602Interface to device under test

Definitions

  • This invention relates generally to semiconductor wafer testing and more particularly to an apparatus and method for reducing the pin count necessary to test Rambus dynamic random access memory (RDRAM).
  • RDRAM Rambus dynamic random access memory
  • Rambus DRAM is a general-purpose, high-performance, packet-oriented dynamic random-access memory (DRAM) device suitable for use in a broad range of applications, including computer memory, graphics, video, and other applications.
  • FIG. 1 schematically illustrates an RDRAM device 10 interconnected with a central processing unit (CPU) 11 as part of a typical computer system.
  • the RDRAM device 10 receives clock signals 12 , control logic signal 14 and address information 16 from the CPU 11 via a controller 20 .
  • Data 17 is written to and read from the RDRAM 10 .
  • FIG. 2 is a block diagram illustrating one 144 Mbit RDRAM configuration in the normal mode.
  • the RDRAM comprises two major blocks: a “core” block 18 comprising banks 22 , sense amps 24 and I/O gating 26 similar to those found in other types of DRAM devices, and a control logic block in normal mode 19 which permits an external controller 20 to access the core 18 .
  • the RDRAM core 18 is internally configured as 32 banks 22 . Each bank 22 has 32,768 144-bit storage locations.
  • FIG. 3 is a diagram indicating that each of the banks 22 is organized as 512 rows 28 by 64 columns 30 by 144 bits 32 .
  • the 144 bits 32 in each column 30 are serially multiplexed onto the RDRAM's I/O pins as eight 18-bit words 34 .
  • the most significant bits 17 - 9 are communicated on I/O pins DQA ⁇ 8:0>, and the least significant bits 8 - 0 are communicated on the I/O pins DQB ⁇ 8:0>.
  • the nine bits on each set of pins are output or input on successive clock edges so that the bits in the eight words are transferred on eight clock edges.
  • the control logic block 19 in FIG. 2 receives the CMD, SCK, SIO 0 , and SIO 1 strobes that supply the RDRAM configuration information to the controller 10 , and that select the operating modes of the RDRAM device 10 .
  • the CFM, CFMN, CTM and CTMN pins generate the internal clocks used to transmit read data, receive write data, and receive the row and column pins used to manage the transfer of data between the banks 22 and the sense amps 24 of the RDRAM 10 .
  • Address information 16 is passed to the RDRAM device 10 from the CPU 11 via eight RQ pins 36 as illustrated in FIG. 4.
  • the RQ pins 36 are divided into two groups.
  • Three ROW pins 38 are de-multiplexed into row packets 40 that manage the transfer of data between the banks 22 and the sense amps 24 .
  • Five COL pins 42 are de-multiplexed into column packets 44 and manage the transfer of data between the data pins and the sense amps 24 of the RDRAM 10 . More detailed information on the operation of RDRAM can be found in Reference A, Direct RDRAM Preliminary Information, Document DL0059 Version 0.9 by Rambus Inc. which is incorporated herein by reference.
  • FIGS. 5A and 5B illustrate a semiconductor chip with exposed bonding pads 46 made of metal, such as aluminum or the like that are formed as terminals of integrated circuits.
  • the control signals 14 , the address signals 16 , and the data 17 are exchanged with the CPU 11 through connections at these bonding pads 46 .
  • FIG. 6 illustrates the semiconductor wafer 48 prior to being diced into individual semiconductor chips.
  • FIG. 6 only shows a relatively small number of semiconductor chips on the wafer, one skilled in the art will appreciate that many semiconductor chips can be cut from a single wafer.
  • the semiconductor chips 10 are subjected to electrical characteristic tests while they are on the wafer 48 through the use of a testing apparatus, e.g., a wafer probe 50 having a plurality of pins 52 . Note that only the head of the wafer probe 50 is shown in FIG. 6.
  • Wafer probe testing is commonly used to quality sort individual semiconductor chips before they are diced from the wafer 48 .
  • the primary goal of wafer probe testing is to identify and mark for easy discrimination defective chips early in the manufacturing process. Wafer testing significantly improves manufacturing efficiency and product quality by detecting defects at the earliest possible stages in the manufacturing and assembly process. In some circumstances, wafer probe testing provides information to enable certain defects to be corrected.
  • FIG. 7 shows a plurality of the conductive pins 52 of the wafer probe 50 of FIG. 6.
  • the pins have respective tip ends 54 positionally adjusted to align with the bonding pads 48 of the RDRAM device 10 to be tested.
  • a wafer probe 50 has a limited number of pins 52 (e.g., 100 pins) available to supply the test signals to the RDRAM device 10 in the wafer 48 .
  • the RDRAM devices 10 could be tested in their normal mode, but this would require in excess of 40 pins 52 on the wafer probe 50 to test each chip 10 .
  • Others have recognized the benefits of creating a special test mode that enables a semiconductor chip such as an RDRAM device 10 to be tested with fewer pins.
  • the prior art method of wafer testing RDRAM chips requires 34 pins 52 to test each RDRAM device 10 , of which 18 pins are address and data pins.
  • the first operation in selecting the address on the RDRAM core entails precharging the bank 22 .
  • Precharging is necessary because adjacent banks 22 share the same sense amps 24 and cannot, therefore be simultaneously activated.
  • Precharging a particular bank 22 deactivates the particular bank and prepares that bank 22 and the sense amps 24 for subsequent activation. For example, when the row 28 in the particular bank 22 is activated, the two adjacent sense amps 24 are connected to or associated with that bank 22 , and therefore are not available for use by the two adjacent banks.
  • Precharging the bank 22 also automatically causes the two adjacent banks to be precharged, thereby ensuring that adjacent banks are not activated at the same time.
  • Selecting one of the 32 banks 22 to precharge requires five address bits to specify the bank address. These address bits are provided in a first control signal.
  • the next operation in selecting an address is activating a row 28 in a selected bank using a second control signal. This operation requires nine address bits to select one of the 512 rows 28 , and five address bits to select one of the 32 banks 22 , for a total of 14 address bits.
  • the next operation reads a column 30 in an open bank using a third control signal. This operation requires five bank bits. This operation also requires six column bits to select one of the 64 columns 30 .
  • the invention comprises a method of testing computer memory devices, such as Rambus DRAM.
  • the method requires fewer pin connections to test each chip on a semiconductor wafer than previously known methods.
  • the test is performed on a semiconductor wafer using a wafer probe.
  • the number of pins required is reduced by using a trailing edge of a precharge clock to latch the bank address, thus eliminating the need to perform this function on a later step.
  • the number of pins required is further reduced by dividing the chip to be tested into a plurality of array cores and compressing the output data so that only one data pin per array core is required.
  • By reducing the pin count more DRAMs can be tested at the same time, thus reducing the overall test cost and time for testing a complete wafer.
  • FIG. 1 is a block diagram illustrating a RDRAM device as part of a computer system.
  • FIG. 2 is a functional block diagram illustrating the RDRAM chip configuration in the normal mode.
  • FIG. 3 is a conceptual drawing illustrating the RDRAM bank configured in rows, columns, words, and bits in the normal mode.
  • FIG. 4 is a conceptual drawing illustrating RQ pins delivering the address information of FIGS. 1 and 2.
  • FIG. 5A is a top plan view of a RDRAM chip illustrating the bonding pads.
  • FIG. 5B is a side elevation of the RDRAM of FIG. 5A.
  • FIG. 6 is a perspective view of a RDRAM semiconductor wafer comprising a plurality of chips with a wafer probe.
  • FIG. 7 is a top plan view of the bonding pads of a RDRAM chip aligned with the conductive pins which are connected to a wafer probe.
  • FIG. 8 is a functional block diagram illustrating the RDRAM chip configuration in the DFT mode.
  • FIGS. 9A and 9B are conceptual drawings, illustrating the RDRAM bank configured in rows, columns, words, and bits and being further divided so that the data from two rows can be compressed for 2 ⁇ row compression and output compressed into a single DQ for DQ compression.
  • FIG. 10 is a block diagram illustrating the RDRAM core divided up into four quadrants with a single DQ output after DQ compression.
  • FIG. 11 is a timing diagram illustrating a typical write cycle in the DFT mode.
  • FIG. 12 is a timing diagram illustrating a typical read cycle in the DFT mode.
  • FIG. 13 is a timing diagram illustrating the compressed data output for a DQ in a window manner showing a fault detection.
  • the RDRAM in accordance with the invention has two modes of operation: (1) a high speed packet mode for normal operation; and (2) a low speed asynchronous mode for testing; which bypasses the packetizing hardware, often called “design for test” circuits or DFT.
  • This second mode shown as a block diagram in FIG. 8 is realized by including DFT mode control logic 58 and data compression logic 59 in the RDRAM device 10 to facilitate testing.
  • the RDRAM in the DFT test mode, behaves similar to an asynchronous DRAM, although data is still input/output in bursts of eight.
  • the RDRAM comprises three major blocks: a “core” block 18 , the control logic block in DFT mode 58 and the Data Compression/Expansion Logic box 59 .
  • the core 18 is internally configured as 32 banks 22 organized as 512 rows 28 by 64 columns 30 by 144-bit storage locations. The 144 bits are multiplexed as eight 18-bit words. The core is further divided for testing purposes as will be discussed below.
  • the DFT control logic 58 receives a number of signals from the wafer probe 50 , including, TestBSENSE, TestPRECH, TestWRITE, TestCOLLAT, TestCLK_R/W, SIO 0 , SIO 1 , CMD, SCK and Burn PRECH_EN.
  • the Data Compression/Expansion Logic 59 compresses data so that only four data pins are required, as will be discussed below.
  • the pins required for the DFT mode of operation are a subset of the pins used in the normal mode of operation. Many of the functions of the normal mode pins are redefined (as discussed below) for the DFT mode. The mapping of the normal mode pins to the DFT mode functions is illustrated below in Table 1.
  • the location must be referenced by its bank address, row address, and column address.
  • selecting the bank address of one of the 32 banks requires five address bits
  • selecting a row address of one of the 512 rows in a bank requires nine address bits
  • selecting a column address of one of the 64 columns in a bank requires six address bits.
  • the 144 Mbit RDRAM device is wafer tested using DQ compression and 2 ⁇ row compression.
  • a 288 Mbit RDRAM device can be tested according to the invention as well.
  • the RDRAM core block 18 is internally configured as 32 banks 22 .
  • Each bank 22 is organized as 512 rows 28 by 128 columns 30 by 144 bits 32 . Selecting the bank address of one of the 32 banks requires five address bits, selecting a row address of one of the 512 rows in a bank requires nine address bits, and selecting a column address of one of the 128 columns in a bank requires seven address bits.
  • the 288 Mbit RDRAM device can be wafer tested using either DQ compression or DQ compression and 2 ⁇ row compression.
  • the RDRAM device 10 is divided into four quadrants, 60 A, 60 B, 60 C, and 60 D, as illustrated in FIG. 10, with each quadrant corresponding to a respective 36 megabit array core 61 A, 61 B, 61 C, and 61 D.
  • Each array core 61 A, 61 B, 61 C, and 61 D is an independent repair region.
  • the lower two quadrants, 60 A and 60 B comprise banks 0 - 15 .
  • the upper two quadrants, 60 C and 60 D comprise banks 16 - 31 . This division is based on physical design parameters of the RDRAM device 10 .
  • the lower left quadrant 60 A comprises bits 9 - 17 of banks 0 - 15 .
  • the lower right quadrant 60 B comprises bits 0 - 8 of banks 0 - 15 .
  • the upper left quadrant 60 C comprises bits 9 - 17 of banks 16 - 31 .
  • the upper right quadrant 60 D comprises bits 0 - 8 of banks 16 - 31 .
  • a data bit DQ 0 is used to test the upper left quadrant 60 C.
  • a data bit DQ 1 is used to test the upper right quadrant 60 D.
  • a data bit DQ 2 is used to test the lower left quadrant 60 A.
  • a data bit DQ 3 is used to test the lower right quadrant 60 B.
  • the upper banks ( 16 - 31 ) and the lower banks ( 0 - 15 ) have separate data connections in the DFT mode. Thus, the most significant bank bit that distinguishes the upper and lower sets of banks is not required, and the number of bank bits is reduced from five bits to four bits.
  • the 2 ⁇ row compression further reduces the number of bank address bits required.
  • the data from corresponding rows in two alternating banks e.g., bank n with bank n+2 and bank n+16 with bank n+18) are combined as shown in FIGS. 9A and 9B so that the data are transferred to and from both rows using a common DQ bit.
  • only three bank bits are required to select one of the eight banks in each quadrant.
  • the data from the two rows of the alternating banks are transferred (either written to the memory or read from the memory) one byte at a time, as in the normal mode.
  • the nine bits of data from each of the two rows (18 bits of data in all) in each quadrant are combined into a respective single bit (i.e., DQ 0 , DQ 1 , DQ 2 , or DQ 3 ).
  • DQ 0 , DQ 1 , DQ 2 , or DQ 3 the data from a column in the two rows are output as a sequence of eight single data bits.
  • the compression of the data bits is performed by the data compression/expansion logic 59 .
  • Each quadrant 60 A, 60 B, 60 C, and 60 D can have an associated data compression/expansion logic 59 A, 59 B, 59 C, and 59 D as illustrated in FIGS. 9A and 9B.
  • Data are written to the memory by applying a data bit to each of the compressed data pins (i.e., to DQ 0 , DQ 1 , DQ 2 , DQ 3 ).
  • the data compression/expansion logic 59 fans out the single data bit to the eighteen data locations addressed by the bank, row and column bits. Thus, the same data are written into all eighteen locations.
  • the data from the eighteen locations read during each clock edge are compared to determine if any location has a different data output. If the data are the same, the output on the DQ line has a first constant state (e.g., a logic one or a logic zero in accordance with the data written during the write operation) to indicate pass. If any bit in any of the eighteen locations is different, the data output on the DQ line is forced to have a transition to indicate a failure.
  • a first constant state e.g., a logic one or a logic zero in accordance with the data written during the write operation
  • the result of the DQ compression and the 2 ⁇ row compression is that the array cores 61 A, 61 B, 61 C and 61 D are configured as 8 banks by 512 rows by 128 columns by eight four-bit bytes. Therefore, only three bank select bits, nine row address bits, and seven column address bits are required to identify a particular location in the array core. This results in the ability to test each RDRAM device 10 using only nine pins on the wafer probe 50 for defining a specific address location. When the row is activated, nine row address bits identify one of the 512 rows. When a column in an open bank is read, the seven column bits identify the column in the bank to be written to or read from.
  • FIG. 11 is a timing diagram that illustrates a typical write cycle that is used to select the bank for row access and the bank for column access, row address, column address, and strobe in the data.
  • FIG. 12 is a timing diagram that similarly illustrates a typical read cycle.
  • address pins 64 , 68 , and 70 refer to subdivisions of the nine address pins used to identify a particular location in the array core.
  • Address pins 64 represent Addr ⁇ 8:6> (three address pins 8 , 7 , and 6 ).
  • Address pins 68 represent Addr ⁇ 5:1> (five address pins 5 , 4 , 3 , 2 , and 1 ).
  • Address pins 70 represent Addr ⁇ 0> (one address pin 0 ).
  • a precharge clock is used to select the bank address.
  • the leading edge of TestPRECH 62 is used to precharge the bank designated by the bank address present on the address pins 64 .
  • Precharging the bank prepares the bank and the sense amps for activation. Since adjacent inner banks share the same sense amps, adjacent banks cannot be activated at the same time. Precharging any bank automatically causes adjacent banks to be precharged also, thereby ensuring that adjacent banks are not open at the same time. This happens in all modes of operation, not just the DFT mode.
  • TestPRECH 62 On the falling edge of TestPRECH 62 , the bank corresponding to the bank address on the address pins 64 is latched. This latched bank address represents the bank that will be activated the next time TestBSENSE is presented. Multiple banks can be active at any one time. That is, banks previously activated and not subsequently deactivated by precharging remain active in addition to the newly activated bank. Precharging banks and latching banks are accomplished using different edges of the same TestPRECH signal 62 . Thus, the present invention eliminates the need to provide separate control signals for the precharge function and the latching function.
  • TestBSENSE 66 causes the selected row of the latched (i.e., active) bank to be sensed.
  • the row address to be sensed is the address present on the address pins 64 , 68 and 70 at the falling edge of TestBSENSE 66 . Because there are 512 rows, nine address pins are required to select the row to be tested. Because the bank was latched using the other edge of the TestPRECH 62 , it is not required to select a bank in this operation. Thus, unlike other known methods, the bank select bits do not have to be applied at this time and only the nine address bits need to be applied.
  • Data are then either read from or written to the column in accordance with the address present on the address pins at the rising edge of a column latch clock, TestCOLLAT 72 .
  • the row address of the bank to be opened is presented on the falling edge of TestBSENSE 66 .
  • the address of the column to be accessed is presented on the rising edge of TestCOLLAT 72 .
  • the address of that bank must be the same as the bank of the column to be accessed. As a result, nine address bits are sufficient to provide the necessary address bits to identify any location in the array core.
  • the bank must be one of the banks that was active when TestBSENSE 66 was applied.
  • TestClkR/W 76 Eight clock edges transfer the data. It takes a total of six TestClkR/W 76 cycles to completely empty the read buffer. The chip under test remains in the output mode until all data are read out of the read buffer. Any additional clock cycles initiates a new read sequence. Note that any transition on TestClkR/W 76 initiates a read or write sequence depending on the state of TestWrite 74 .
  • FIG. 13 is a timing diagram that illustrates the compressed data being output in a window manner when reading the compressed DQs. If the expected data is a “0”, then the DQ will be low during the entire window. A failure is indicated if the wrong data is present, or if a data transition is detected during the window. If the expected data is “1”, then the DQ should remain high throughout the window.

Abstract

An apparatus and a method are disclosed for reducing the pin driver count required for testing computer memory devices, specifically Rambus DRAM, while a die is on a semiconductor wafer. By reducing the pin count, more DRAMs can be tested at the same time, thereby reducing test cost and time. One preferred embodiment utilizes a trailing edge of a precharge clock to select a new active bank address, so that the address line required to select a new active address does not have to be accessed at the same time as the row lines.

Description

    RELATED APPLICATIONS
  • This present application is a continuation of U.S. patent application Ser. No. 09/653,112, filed on Aug. 31, 2000, which is a continuation-in-part application of U.S. patent application Ser. No. 09/454,808, filed on Dec. 3, 1999, now U.S. Pat. No. 6,530,045, issued on Mar. 4, 2003.[0001]
  • BACKGROUND OF THE INVENTION
  • 1. Field of the Invention [0002]
  • This invention relates generally to semiconductor wafer testing and more particularly to an apparatus and method for reducing the pin count necessary to test Rambus dynamic random access memory (RDRAM). [0003]
  • 2. Description of the Related Art [0004]
  • Rambus DRAM (RDRAM) is a general-purpose, high-performance, packet-oriented dynamic random-access memory (DRAM) device suitable for use in a broad range of applications, including computer memory, graphics, video, and other applications. FIG. 1 schematically illustrates an [0005] RDRAM device 10 interconnected with a central processing unit (CPU) 11 as part of a typical computer system. The RDRAM device 10 receives clock signals 12, control logic signal 14 and address information 16 from the CPU 11 via a controller 20. Data 17 is written to and read from the RDRAM 10.
  • FIG. 2 is a block diagram illustrating one 144 Mbit RDRAM configuration in the normal mode. The RDRAM comprises two major blocks: a “core” [0006] block 18 comprising banks 22, sense amps 24 and I/O gating 26 similar to those found in other types of DRAM devices, and a control logic block in normal mode 19 which permits an external controller 20 to access the core 18. The RDRAM core 18 is internally configured as 32 banks 22. Each bank 22 has 32,768 144-bit storage locations.
  • FIG. 3 is a diagram indicating that each of the [0007] banks 22 is organized as 512 rows 28 by 64 columns 30 by 144 bits 32. The 144 bits 32 in each column 30 are serially multiplexed onto the RDRAM's I/O pins as eight 18-bit words 34. The most significant bits 17-9 are communicated on I/O pins DQA <8:0>, and the least significant bits 8-0 are communicated on the I/O pins DQB <8:0>. The nine bits on each set of pins are output or input on successive clock edges so that the bits in the eight words are transferred on eight clock edges.
  • The [0008] control logic block 19 in FIG. 2 receives the CMD, SCK, SIO0, and SIO1 strobes that supply the RDRAM configuration information to the controller 10, and that select the operating modes of the RDRAM device 10. The CFM, CFMN, CTM and CTMN pins generate the internal clocks used to transmit read data, receive write data, and receive the row and column pins used to manage the transfer of data between the banks 22 and the sense amps 24 of the RDRAM 10.
  • [0009] Address information 16 is passed to the RDRAM device 10 from the CPU 11 via eight RQ pins 36 as illustrated in FIG. 4. The RQ pins 36 are divided into two groups. Three ROW pins 38 are de-multiplexed into row packets 40 that manage the transfer of data between the banks 22 and the sense amps 24. Five COL pins 42 are de-multiplexed into column packets 44 and manage the transfer of data between the data pins and the sense amps 24 of the RDRAM 10. More detailed information on the operation of RDRAM can be found in Reference A, Direct RDRAM Preliminary Information, Document DL0059 Version 0.9 by Rambus Inc. which is incorporated herein by reference.
  • Semiconductor chips, such as an [0010] RDRAM device 10, contain circuit elements formed in the semiconductor layers which make up the integrated circuits. FIGS. 5A and 5B illustrate a semiconductor chip with exposed bonding pads 46 made of metal, such as aluminum or the like that are formed as terminals of integrated circuits. In normal operation, the control signals 14, the address signals 16, and the data 17 are exchanged with the CPU 11 through connections at these bonding pads 46.
  • In the manufacturing process, a large number of semiconductor chips, each having a predetermined circuit pattern, are formed on a [0011] semiconductor wafer 48 such as that shown in FIG. 6. FIG. 6 illustrates the semiconductor wafer 48 prior to being diced into individual semiconductor chips. Although FIG. 6 only shows a relatively small number of semiconductor chips on the wafer, one skilled in the art will appreciate that many semiconductor chips can be cut from a single wafer. The semiconductor chips 10 are subjected to electrical characteristic tests while they are on the wafer 48 through the use of a testing apparatus, e.g., a wafer probe 50 having a plurality of pins 52. Note that only the head of the wafer probe 50 is shown in FIG. 6. Wafer probe testing is commonly used to quality sort individual semiconductor chips before they are diced from the wafer 48. The primary goal of wafer probe testing is to identify and mark for easy discrimination defective chips early in the manufacturing process. Wafer testing significantly improves manufacturing efficiency and product quality by detecting defects at the earliest possible stages in the manufacturing and assembly process. In some circumstances, wafer probe testing provides information to enable certain defects to be corrected.
  • FIG. 7 shows a plurality of the [0012] conductive pins 52 of the wafer probe 50 of FIG. 6. The pins have respective tip ends 54 positionally adjusted to align with the bonding pads 48 of the RDRAM device 10 to be tested. A wafer probe 50 has a limited number of pins 52 (e.g., 100 pins) available to supply the test signals to the RDRAM device 10 in the wafer 48. The RDRAM devices 10 could be tested in their normal mode, but this would require in excess of 40 pins 52 on the wafer probe 50 to test each chip 10. Others have recognized the benefits of creating a special test mode that enables a semiconductor chip such as an RDRAM device 10 to be tested with fewer pins. Therefore, one skilled in the art will recognize that it is not required to have a pin 52 for every bonding pad 48 on the chip 10. However, prior testing methodology for RDRAM devices 10 requires at least 34 pins 52 on the wafer probe 50 to test each RDRAM device 10. Consequently, the 100 pin wafer probe is restricted to test, at most, two semiconductor chips at one time. As a result, the production time and chip costs are negatively impacted by this limitation.
  • As set forth above, the prior art method of wafer testing RDRAM chips requires 34 [0013] pins 52 to test each RDRAM device 10, of which 18 pins are address and data pins. Following this method, the first operation in selecting the address on the RDRAM core entails precharging the bank 22. Precharging is necessary because adjacent banks 22 share the same sense amps 24 and cannot, therefore be simultaneously activated. Precharging a particular bank 22 deactivates the particular bank and prepares that bank 22 and the sense amps 24 for subsequent activation. For example, when the row 28 in the particular bank 22 is activated, the two adjacent sense amps 24 are connected to or associated with that bank 22, and therefore are not available for use by the two adjacent banks. Precharging the bank 22 also automatically causes the two adjacent banks to be precharged, thereby ensuring that adjacent banks are not activated at the same time.
  • Selecting one of the 32 [0014] banks 22 to precharge requires five address bits to specify the bank address. These address bits are provided in a first control signal. The next operation in selecting an address is activating a row 28 in a selected bank using a second control signal. This operation requires nine address bits to select one of the 512 rows 28, and five address bits to select one of the 32 banks 22, for a total of 14 address bits. The next operation reads a column 30 in an open bank using a third control signal. This operation requires five bank bits. This operation also requires six column bits to select one of the 64 columns 30.
  • Reducing the number of address bits required to specify the address location to be tested reduces the number of [0015] pin connections 52 required on the wafer probe 50 to test each individual RDRAM device 10. Reducing the required number of pin connections 52 therefore allows more devices 10 to be tested at the same time, thus permitting an important reduction in production time and chip costs. As chip sizes continue to decrease, there is a corresponding increase in the number of chips on each semiconductor wafer to be tested. Therefore, the ability to test an increased number of devices at the same time grows in importance.
  • SUMMARY OF THE INVENTION
  • The invention comprises a method of testing computer memory devices, such as Rambus DRAM. The method requires fewer pin connections to test each chip on a semiconductor wafer than previously known methods. The test is performed on a semiconductor wafer using a wafer probe. The number of pins required is reduced by using a trailing edge of a precharge clock to latch the bank address, thus eliminating the need to perform this function on a later step. In combination with such use of the precharge clock's trailing edge, the number of pins required is further reduced by dividing the chip to be tested into a plurality of array cores and compressing the output data so that only one data pin per array core is required. By reducing the pin count, more DRAMs can be tested at the same time, thus reducing the overall test cost and time for testing a complete wafer.[0016]
  • BRIEF DESCRIPTION OF THE DRAWINGS
  • FIG. 1 is a block diagram illustrating a RDRAM device as part of a computer system. [0017]
  • FIG. 2 is a functional block diagram illustrating the RDRAM chip configuration in the normal mode. [0018]
  • FIG. 3 is a conceptual drawing illustrating the RDRAM bank configured in rows, columns, words, and bits in the normal mode. [0019]
  • FIG. 4 is a conceptual drawing illustrating RQ pins delivering the address information of FIGS. 1 and 2. [0020]
  • FIG. 5A is a top plan view of a RDRAM chip illustrating the bonding pads. [0021]
  • FIG. 5B is a side elevation of the RDRAM of FIG. 5A. [0022]
  • FIG. 6 is a perspective view of a RDRAM semiconductor wafer comprising a plurality of chips with a wafer probe. [0023]
  • FIG. 7 is a top plan view of the bonding pads of a RDRAM chip aligned with the conductive pins which are connected to a wafer probe. [0024]
  • FIG. 8 is a functional block diagram illustrating the RDRAM chip configuration in the DFT mode. [0025]
  • FIGS. 9A and 9B are conceptual drawings, illustrating the RDRAM bank configured in rows, columns, words, and bits and being further divided so that the data from two rows can be compressed for 2× row compression and output compressed into a single DQ for DQ compression. [0026]
  • FIG. 10 is a block diagram illustrating the RDRAM core divided up into four quadrants with a single DQ output after DQ compression. [0027]
  • FIG. 11 is a timing diagram illustrating a typical write cycle in the DFT mode. [0028]
  • FIG. 12 is a timing diagram illustrating a typical read cycle in the DFT mode. [0029]
  • FIG. 13 is a timing diagram illustrating the compressed data output for a DQ in a window manner showing a fault detection.[0030]
  • DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
  • The RDRAM in accordance with the invention has two modes of operation: (1) a high speed packet mode for normal operation; and (2) a low speed asynchronous mode for testing; which bypasses the packetizing hardware, often called “design for test” circuits or DFT. This second mode, shown as a block diagram in FIG. 8 is realized by including DFT [0031] mode control logic 58 and data compression logic 59 in the RDRAM device 10 to facilitate testing. In one embodiment of the invention, in the DFT test mode, the RDRAM behaves similar to an asynchronous DRAM, although data is still input/output in bursts of eight.
  • As shown in FIG. 8, the RDRAM comprises three major blocks: a “core” [0032] block 18, the control logic block in DFT mode 58 and the Data Compression/Expansion Logic box 59. As shown in FIGS. 9A and 9B, the core 18 is internally configured as 32 banks 22 organized as 512 rows 28 by 64 columns 30 by 144-bit storage locations. The 144 bits are multiplexed as eight 18-bit words. The core is further divided for testing purposes as will be discussed below.
  • The [0033] DFT control logic 58 receives a number of signals from the wafer probe 50, including, TestBSENSE, TestPRECH, TestWRITE, TestCOLLAT, TestCLK_R/W, SIO0, SIO1, CMD, SCK and Burn PRECH_EN. The Data Compression/Expansion Logic 59 compresses data so that only four data pins are required, as will be discussed below.
  • The pins required for the DFT mode of operation are a subset of the pins used in the normal mode of operation. Many of the functions of the normal mode pins are redefined (as discussed below) for the DFT mode. The mapping of the normal mode pins to the DFT mode functions is illustrated below in Table 1. [0034]
    TABLE 1
    DFT Pin Mapping
    Pin DFT Function
    SCK SCK
    CMD CMD
    SIO<1:0> SIO<1:0>
    CFM/CTM TestClkR/W
    RQ<0> TestBSENSE
    RQ<1> TestPRECH
    RQ<2> TestWrite
    RQ<3> TestCOLLAT
    DQB<2:0> ADR<2:0>
    DQA<3:0> ADR<6:3>
    DQB<3> ADR<7>
    DQB<6> ADR<8>
    DQB<8> Burn PRECH_EN
    DQA<5:4> DQ<1:0>
    DQB<5:4> DQ<3:2>
    CFMN/CTMN VCC/2
    VCMOS VCMOS
  • To test a specific location in the [0035] core block 18 of the RDRAM device 10, the location must be referenced by its bank address, row address, and column address. In the normal configuration of a 144 Mbit RDRAM device as illustrated in FIG. 3, selecting the bank address of one of the 32 banks requires five address bits, selecting a row address of one of the 512 rows in a bank requires nine address bits, and selecting a column address of one of the 64 columns in a bank requires six address bits. In accordance with the present invention, the 144 Mbit RDRAM device is wafer tested using DQ compression and 2× row compression.
  • In a further embodiment, a 288 Mbit RDRAM device can be tested according to the invention as well. In the normal configuration of a 288 Mbit RDRAM device, the [0036] RDRAM core block 18 is internally configured as 32 banks 22. Each bank 22 is organized as 512 rows 28 by 128 columns 30 by 144 bits 32. Selecting the bank address of one of the 32 banks requires five address bits, selecting a row address of one of the 512 rows in a bank requires nine address bits, and selecting a column address of one of the 128 columns in a bank requires seven address bits. In accordance with the present invention, the 288 Mbit RDRAM device can be wafer tested using either DQ compression or DQ compression and 2× row compression.
  • In DQ compression, the [0037] RDRAM device 10 is divided into four quadrants, 60A, 60B, 60C, and 60D, as illustrated in FIG. 10, with each quadrant corresponding to a respective 36 megabit array core 61A, 61B, 61C, and 61D. Each array core 61A, 61B, 61C, and 61D is an independent repair region. The lower two quadrants, 60A and 60B, comprise banks 0-15. The upper two quadrants, 60C and 60D, comprise banks 16-31. This division is based on physical design parameters of the RDRAM device 10. The lower left quadrant 60A comprises bits 9-17 of banks 0-15. The lower right quadrant 60B comprises bits 0-8 of banks 0-15. The upper left quadrant 60C comprises bits 9-17 of banks 16-31. The upper right quadrant 60D comprises bits 0-8 of banks 16-31. As discussed below, for testing, only a single bit of data is transferred into and out of each quadrant 60A, 60B, 60C, and 60D. In particular, as will be discussed below, a data bit DQ0 is used to test the upper left quadrant 60C. A data bit DQ1 is used to test the upper right quadrant 60D. A data bit DQ2 is used to test the lower left quadrant 60A. A data bit DQ3 is used to test the lower right quadrant 60B. Therefore, only four data bits are required to test the entire memory. Note further that the upper banks (16-31) and the lower banks (0-15) have separate data connections in the DFT mode. Thus, the most significant bank bit that distinguishes the upper and lower sets of banks is not required, and the number of bank bits is reduced from five bits to four bits.
  • In one embodiment of the invention using DQ compression and 2× row compression, the 2× row compression further reduces the number of bank address bits required. In particular, the data from corresponding rows in two alternating banks (e.g., bank n with bank n+2 and bank n+16 with bank n+18) are combined as shown in FIGS. 9A and 9B so that the data are transferred to and from both rows using a common DQ bit. This reduces the number of selectable banks in each quadrant from sixteen to eight. Thus, only three bank bits are required to select one of the eight banks in each quadrant. [0038]
  • The data from the two rows of the alternating banks are transferred (either written to the memory or read from the memory) one byte at a time, as in the normal mode. However, because only one data pin is available for each [0039] quadrant 60A, 60B, 60C, and 60D, the nine bits of data from each of the two rows (18 bits of data in all) in each quadrant are combined into a respective single bit (i.e., DQ0, DQ1, DQ2, or DQ3). Thus, for each quadrant the data from a column in the two rows are output as a sequence of eight single data bits.
  • The compression of the data bits is performed by the data compression/[0040] expansion logic 59. Each quadrant 60A, 60B, 60C, and 60D can have an associated data compression/ expansion logic 59A, 59B, 59C, and 59D as illustrated in FIGS. 9A and 9B. Data are written to the memory by applying a data bit to each of the compressed data pins (i.e., to DQ0, DQ1, DQ2, DQ3). On each clock edge the data compression/expansion logic 59 fans out the single data bit to the eighteen data locations addressed by the bank, row and column bits. Thus, the same data are written into all eighteen locations. Thereafter, when the memory locations are read to test the integrity of the memory, the data from the eighteen locations read during each clock edge are compared to determine if any location has a different data output. If the data are the same, the output on the DQ line has a first constant state (e.g., a logic one or a logic zero in accordance with the data written during the write operation) to indicate pass. If any bit in any of the eighteen locations is different, the data output on the DQ line is forced to have a transition to indicate a failure.
  • In one embodiment for testing a 288 Mbit RDRAM device, the result of the DQ compression and the 2× row compression is that the [0041] array cores 61A, 61B, 61C and 61D are configured as 8 banks by 512 rows by 128 columns by eight four-bit bytes. Therefore, only three bank select bits, nine row address bits, and seven column address bits are required to identify a particular location in the array core. This results in the ability to test each RDRAM device 10 using only nine pins on the wafer probe 50 for defining a specific address location. When the row is activated, nine row address bits identify one of the 512 rows. When a column in an open bank is read, the seven column bits identify the column in the bank to be written to or read from.
  • FIG. 11 is a timing diagram that illustrates a typical write cycle that is used to select the bank for row access and the bank for column access, row address, column address, and strobe in the data. FIG. 12 is a timing diagram that similarly illustrates a typical read cycle. In FIGS. 11 and 12, address pins [0042] 64, 68, and 70 refer to subdivisions of the nine address pins used to identify a particular location in the array core. Address pins 64 represent Addr<8:6> (three address pins 8, 7, and 6). Address pins 68 represent Addr<5:1> (five address pins 5, 4, 3, 2, and 1). Address pins 70 represent Addr<0> (one address pin 0).
  • In the write and read cycles depicted in FIGS. 11 and 12, respectively, a precharge clock, [0043] TestPRECH 62, is used to select the bank address. The leading edge of TestPRECH 62 is used to precharge the bank designated by the bank address present on the address pins 64. Precharging the bank prepares the bank and the sense amps for activation. Since adjacent inner banks share the same sense amps, adjacent banks cannot be activated at the same time. Precharging any bank automatically causes adjacent banks to be precharged also, thereby ensuring that adjacent banks are not open at the same time. This happens in all modes of operation, not just the DFT mode.
  • On the falling edge of [0044] TestPRECH 62, the bank corresponding to the bank address on the address pins 64 is latched. This latched bank address represents the bank that will be activated the next time TestBSENSE is presented. Multiple banks can be active at any one time. That is, banks previously activated and not subsequently deactivated by precharging remain active in addition to the newly activated bank. Precharging banks and latching banks are accomplished using different edges of the same TestPRECH signal 62. Thus, the present invention eliminates the need to provide separate control signals for the precharge function and the latching function.
  • Next, a row address is selected using address pins and a row sense clock, [0045] TestBSENSE 66. TestBSENSE 66 causes the selected row of the latched (i.e., active) bank to be sensed. The row address to be sensed is the address present on the address pins 64, 68 and 70 at the falling edge of TestBSENSE 66. Because there are 512 rows, nine address pins are required to select the row to be tested. Because the bank was latched using the other edge of the TestPRECH 62, it is not required to select a bank in this operation. Thus, unlike other known methods, the bank select bits do not have to be applied at this time and only the nine address bits need to be applied.
  • Data are then either read from or written to the column in accordance with the address present on the address pins at the rising edge of a column latch clock, TestCOLLAT [0046] 72. The row address of the bank to be opened is presented on the falling edge of TestBSENSE 66. The address of the column to be accessed is presented on the rising edge of TestCOLLAT 72. In one embodiment of the invention, if a new bank is to be opened, then the address of that bank must be the same as the bank of the column to be accessed. As a result, nine address bits are sufficient to provide the necessary address bits to identify any location in the array core.
  • In a further embodiment, the bank must be one of the banks that was active when [0047] TestBSENSE 66 was applied. A TestWrite clock 74 determines whether the operation performed at TestCOLLAT 72 time is a read or a write function. If TestWrite=1 at the rising edge of TestCOLLAT 72, then the data present in a write buffer are written to the RDRAM core. If TestWrite=0 at the rising edge of TestCOLLAT 72, then the data are read from the RDRAM core to a read buffer.
  • FIGS. 11 and 12 show a TestClkR/W clock [0048] 76 strobing data into the write buffer or out of the read buffer depending on the state of TestWrite 74. If TestWrite=1, then data are input into the write buffer from the tester on sequential edges of TestClkR/W 76, beginning with the first falling edge. Eight clock edges transfer data. It takes a total of six TesClkR/W 76 cycles to completely load the write buffer. Additional clock cycles will initiate another load sequence. A load sequence is not terminated until the exact number of clock cycles are provided. If TestWrite=0, then data are read from the read buffer to the external bus on each edge of TestClkR/W 76, beginning with the second falling edge. Eight clock edges transfer the data. It takes a total of six TestClkR/W 76 cycles to completely empty the read buffer. The chip under test remains in the output mode until all data are read out of the read buffer. Any additional clock cycles initiates a new read sequence. Note that any transition on TestClkR/W 76 initiates a read or write sequence depending on the state of TestWrite 74.
  • FIG. 13 is a timing diagram that illustrates the compressed data being output in a window manner when reading the compressed DQs. If the expected data is a “0”, then the DQ will be low during the entire window. A failure is indicated if the wrong data is present, or if a data transition is detected during the window. If the expected data is “1”, then the DQ should remain high throughout the window. [0049]
  • If a fault is indicated, it is not necessary to determine which bit failed, it is sufficient to localize the fault to a row. The tester has the capability to reconfigure the chip so that a spare row is used to replace the row with the fault. The technology for such reconfiguration is well known in the field. [0050]
  • Note that by reducing the required address bits to three and by using both edges of the TestPRECH control signal, the maximum number of address bits required is nine, which with the addition of the four data bits, totals thirteen. This is significantly fewer than the eighteen data and address bits used in other known test methods. [0051]
  • Although specific implementations and operation of the invention have been described above with reference to specific embodiments, the invention may be embodied in other forms without departing from the spirit or central characteristics of the invention. The described embodiments are to be considered in all respects only as illustrative and not restrictive. The scope of the invention is indicated by the appended claims rather than by the foregoing description. All changes which come within the meaning of equivalency of the claims are to be embraced within their scope. [0052]

Claims (49)

1. Cancelled.
2. A method of testing a memory device, the method comprising:
precharging a bank of locations of a memory device at a time defined by a first portion of a first control signal;
latching the bank of locations at a time defined by a second portion of the first control signal;
selecting a row address in the bank of locations at a time defined by a first portion of a second control signal;
selecting a column address at a time defined by a first portion of a third control signal; and
outputting data stored at a location of the memory device defined by the selected row address and the selected column address.
3. The method of claim 2, wherein the first portion of the first control signal is a leading edge of the first control signal.
4. The method of claim 3, wherein the second portion of the first control signal is a trailing edge of the first control signal.
5. The method of claim 2, wherein the bank of locations is divided into a plurality of array cores, wherein each array core comprises an independent repair region.
6. The method of claim 2, wherein the memory device comprises a Rambus dynamic random access memory (RDRAM).
7. The method of claim 2, additionally comprising receiving at most four address bits that identify the bank of locations to be precharged.
8. The method of claim 2, additionally comprising receiving at most three address bits that identify the bank of locations to be precharged.
9. The method of claim 2, wherein the act of selecting the row address comprises receiving nine address bits.
10. The method of claim 2, wherein the act of selecting the column address comprises receiving six address bits.
11. The method of claim 2, wherein the act of selecting the column address comprises receiving seven address bits.
12. A method of testing a memory device, the method comprising:
precharging a bank of locations of a memory device at a time defined by a first portion of a control signal;
latching the bank of locations at a time defined by a second portion of the control signal;
selecting a row address in the bank of locations;
selecting a column address in the bank of locations; and
outputting data stored at a location of the memory device defined by the selected row address and the selected column address.
13. The method of claim 12, wherein the first portion of the control signal is a leading edge of the control signal.
14. The method of claim 13, wherein the second portion of the control signal is a trailing edge of the control signal.
15. The method of claim 12, wherein the bank of locations is divided into a plurality of array cores, wherein each array core comprises an independent repair region.
16. The method of claim 12, wherein the memory device comprises a Rambus dynamic random access memory (RDRAM).
17. The method of claim 12, additionally comprising receiving at most four address bits that identify the bank of locations to be precharged.
18. The method of claim 12, additionally comprising receiving at most three address bits that identify the bank of locations to be precharged.
19. The method of claim 12, wherein the act of selecting the row address comprises receiving nine address bits.
20. The method of claim 12, wherein the act of selecting the column address comprises receiving six address bits.
21. The method of claim 12, wherein the act of selecting the column address comprises receiving seven address bits.
22. A memory device comprising:
a plurality of banks; and
control logic configured to receive from a testing device a first control signal having a precharge signal defined by a first portion of the first control signal and a latch signal defined by a second portion of the first control signal,
and wherein the control logic is further configured to receive a bank address that identifies at least one bank, of the plurality of banks, to be precharged, and wherein the control logic is configured to receive a row address in the at least one precharged bank at a time defined by a second control signal and a column address in the at least one precharged bank at a time defined by a third control signal.
23. The memory device of claim 22, wherein the first portion of the first control signal is a leading edge of the first control signal.
24. The memory device of claim 23, wherein the second portion of the first control signal is a trailing edge of the first control signal.
25. The memory device of claim 22, wherein the control logic is configured to operate in a design for test (DFT) mode.
26. The memory device of claim 22, further comprising data compression logic configured to receive test data from the testing device.
27. The memory device of claim 26, wherein the bank address identifies multiple banks, of the plurality of banks, to be precharged.
28. The memory device of claim 27, wherein the row address and the column address identify locations in the multiple precharged banks.
29. The memory device of claim 28, wherein the data compression logic is configured to output the same test data to each location in the multiple precharged banks.
30. The memory device of claim 22, further comprising a plurality of bonding pads configured to receive the first, second and third control signals from the testing device.
31. The memory device of claim 30, wherein the control logic is configured to receive the first control signal through a single bonding pad.
32. A method of preparing computer memory device location to be tested, the method comprising:
precharging a bank of locations of a computer memory device at a time defined by a first portion of a control signal; and
latching the bank of locations at a time defined by a second portion of the control signal.
33. The method of claim 32, wherein the first portion of the control signal is a leading edge of the control signal.
34. The method of claim 33, wherein the second portion of the control signal is a trailing edge of the control signal.
35. The method of claim 32, wherein the bank of locations is divided into a plurality of array cores, wherein each array core comprises an independent repair region.
36. The method of claim 32, wherein the computer memory device comprises a Rambus dynamic random access memory (RDRAM).
37. A memory device comprising:
a plurality of array cores, each array core comprising a plurality of banks; and
control logic configured to receive a control signal having a precharge signal defined by a first portion of the control signal and a latch signal defined by a second portion of the control signal, and wherein the control logic is further configured to receive a bank address that identifies at least one bank, of the plurality of banks, to be precharged.
38. The memory device of claim 37, wherein the first portion of the control signal is a leading edge of the control signal.
39. The memory device of claim 37, wherein the second portion of the control signal is a trailing edge of the control signal.
40. The memory device of claim 37, wherein the control logic is configured to operate in a design for test (DFT) mode.
41. The memory device of claim 37, further comprising data compression logic configured to receive test data from the testing device.
42. The memory device of claim 41, wherein the bank address identifies multiple banks, of the plurality of banks, to be precharged.
43. The memory device of claim 42, wherein the data compression logic is configured to output the same test data to locations in the multiple precharged banks.
44. The memory device of claim 37, further comprising a plurality of bonding pads.
45. The memory device of claim 44, wherein the control logic is configured to receive the control signal through a single bonding pad.
46. A memory device comprising:
a plurality of array cores, each array core comprising a plurality of banks; and
means for receiving a control signal having a precharge signal defined by a first portion of the control signal and a latch signal defined by a second portion of the control signal, and wherein the means for receiving is further configured to receive a bank address that identifies at least one bank, of the plurality of banks, to be precharged.
47. The memory device of claim 46, wherein the first portion of the control signal is a leading edge of the control signal.
48. The memory device of claim 46, wherein the second portion of the control signal is a trailing edge of the control signal.
49. The memory device of claim 46, wherein the means for receiving is configured to operate in a design for test (DFT) mode.
US10/853,573 1999-12-03 2004-05-25 Apparatus and method for reducing test resources in testing DRAMs Expired - Fee Related US7168018B2 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
US10/853,573 US7168018B2 (en) 1999-12-03 2004-05-25 Apparatus and method for reducing test resources in testing DRAMs
US11/445,944 US20070168790A1 (en) 1999-12-03 2006-06-02 Apparatus and method for reducing test resources in testing drams

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
US09/454,808 US6530045B1 (en) 1999-12-03 1999-12-03 Apparatus and method for testing rambus DRAMs
US09/653,112 US6854079B1 (en) 1999-12-03 2000-08-31 Apparatus and method for reducing test resources in testing Rambus DRAMs
US10/853,573 US7168018B2 (en) 1999-12-03 2004-05-25 Apparatus and method for reducing test resources in testing DRAMs

Related Parent Applications (1)

Application Number Title Priority Date Filing Date
US09/653,112 Continuation US6854079B1 (en) 1999-12-03 2000-08-31 Apparatus and method for reducing test resources in testing Rambus DRAMs

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US11/445,944 Continuation US20070168790A1 (en) 1999-12-03 2006-06-02 Apparatus and method for reducing test resources in testing drams

Publications (2)

Publication Number Publication Date
US20040255211A1 true US20040255211A1 (en) 2004-12-16
US7168018B2 US7168018B2 (en) 2007-01-23

Family

ID=23806198

Family Applications (6)

Application Number Title Priority Date Filing Date
US09/454,808 Expired - Lifetime US6530045B1 (en) 1999-12-03 1999-12-03 Apparatus and method for testing rambus DRAMs
US09/653,112 Expired - Fee Related US6854079B1 (en) 1999-12-03 2000-08-31 Apparatus and method for reducing test resources in testing Rambus DRAMs
US10/853,573 Expired - Fee Related US7168018B2 (en) 1999-12-03 2004-05-25 Apparatus and method for reducing test resources in testing DRAMs
US10/879,437 Expired - Fee Related US6986084B2 (en) 1999-12-03 2004-06-29 Apparatus and method for reducing test resources in testing DRAMS
US11/188,932 Abandoned US20050262405A1 (en) 1999-12-03 2005-07-25 Apparatus and method for reducing test resources in testing DRAMs
US11/445,944 Abandoned US20070168790A1 (en) 1999-12-03 2006-06-02 Apparatus and method for reducing test resources in testing drams

Family Applications Before (2)

Application Number Title Priority Date Filing Date
US09/454,808 Expired - Lifetime US6530045B1 (en) 1999-12-03 1999-12-03 Apparatus and method for testing rambus DRAMs
US09/653,112 Expired - Fee Related US6854079B1 (en) 1999-12-03 2000-08-31 Apparatus and method for reducing test resources in testing Rambus DRAMs

Family Applications After (3)

Application Number Title Priority Date Filing Date
US10/879,437 Expired - Fee Related US6986084B2 (en) 1999-12-03 2004-06-29 Apparatus and method for reducing test resources in testing DRAMS
US11/188,932 Abandoned US20050262405A1 (en) 1999-12-03 2005-07-25 Apparatus and method for reducing test resources in testing DRAMs
US11/445,944 Abandoned US20070168790A1 (en) 1999-12-03 2006-06-02 Apparatus and method for reducing test resources in testing drams

Country Status (1)

Country Link
US (6) US6530045B1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060085703A1 (en) * 2004-10-15 2006-04-20 Chang-Hyuk Lee Memory cell test circuit for use in semiconductor memory device and its method
US20090273991A1 (en) * 2008-04-30 2009-11-05 Seong-Hwi Song Semiconductor memory device, operating method thereof, and compression test method thereof

Families Citing this family (30)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100318266B1 (en) * 1999-06-28 2001-12-24 박종섭 output data compression method and packet command driving type memory device
US6530045B1 (en) * 1999-12-03 2003-03-04 Micron Technology, Inc. Apparatus and method for testing rambus DRAMs
US6590382B2 (en) * 2000-12-22 2003-07-08 International Business Machines Corp. Signal pin tester for AC defects in integrated circuits
KR100557948B1 (en) * 2003-06-20 2006-03-10 주식회사 하이닉스반도체 A method for testing a memory device
US8732644B1 (en) 2003-09-15 2014-05-20 Nvidia Corporation Micro electro mechanical switch system and method for testing and configuring semiconductor functional circuits
US8788996B2 (en) * 2003-09-15 2014-07-22 Nvidia Corporation System and method for configuring semiconductor functional circuits
US8775997B2 (en) * 2003-09-15 2014-07-08 Nvidia Corporation System and method for testing and configuring semiconductor functional circuits
DE10345981B4 (en) * 2003-10-02 2007-10-18 Qimonda Ag A data processing circuit apparatus and method for connecting a circuit core module to an external circuit module
US6880940B1 (en) * 2003-11-10 2005-04-19 Honda Motor Co., Ltd. Magnesium mirror base with countermeasures for galvanic corrosion
US8711161B1 (en) 2003-12-18 2014-04-29 Nvidia Corporation Functional component compensation reconfiguration system and method
US8723231B1 (en) 2004-09-15 2014-05-13 Nvidia Corporation Semiconductor die micro electro-mechanical switch management system and method
US8711156B1 (en) 2004-09-30 2014-04-29 Nvidia Corporation Method and system for remapping processing elements in a pipeline of a graphics processing unit
JP2006120250A (en) * 2004-10-21 2006-05-11 Fujitsu Ltd Semiconductor device and test method therefor
US8021193B1 (en) 2005-04-25 2011-09-20 Nvidia Corporation Controlled impedance display adapter
US7793029B1 (en) 2005-05-17 2010-09-07 Nvidia Corporation Translation device apparatus for configuring printed circuit board connectors
US8417838B2 (en) * 2005-12-12 2013-04-09 Nvidia Corporation System and method for configurable digital communication
US8412872B1 (en) 2005-12-12 2013-04-02 Nvidia Corporation Configurable GPU and method for graphics processing using a configurable GPU
US7292487B1 (en) * 2006-05-10 2007-11-06 Micron Technology, Inc. Independent polling for multi-page programming
US7953510B2 (en) * 2007-09-13 2011-05-31 International Business Machines Corporation System and method for semiconductor identification chip read out
US8724483B2 (en) 2007-10-22 2014-05-13 Nvidia Corporation Loopback configuration for bi-directional interfaces
US20090225610A1 (en) * 2008-03-05 2009-09-10 Wolfgang Hokenmaier Integrated circuit that selectively outputs subsets of a group of data bits
US8687639B2 (en) * 2009-06-04 2014-04-01 Nvidia Corporation Method and system for ordering posted packets and non-posted packets transfer
US9176909B2 (en) 2009-12-11 2015-11-03 Nvidia Corporation Aggregating unoccupied PCI-e links to provide greater bandwidth
US9331869B2 (en) * 2010-03-04 2016-05-03 Nvidia Corporation Input/output request packet handling techniques by a device specific kernel mode driver
KR101188261B1 (en) 2010-07-30 2012-10-05 에스케이하이닉스 주식회사 Multi Bit Test Circuit
KR20120098094A (en) * 2011-02-28 2012-09-05 에스케이하이닉스 주식회사 Semiconductor memory device
KR101282722B1 (en) * 2011-03-09 2013-07-04 에스케이하이닉스 주식회사 Memory device and test method for memory device
US9330031B2 (en) 2011-12-09 2016-05-03 Nvidia Corporation System and method for calibration of serial links using a serial-to-parallel loopback
US8739090B1 (en) * 2013-03-15 2014-05-27 Cadence Design Systems, Inc. Probe signal compression method and apparatus for hardware based verification platforms
CN116092564B (en) * 2023-03-14 2023-06-23 长鑫存储技术有限公司 Memory and testing method

Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5126973A (en) * 1990-02-14 1992-06-30 Texas Instruments Incorporated Redundancy scheme for eliminating defects in a memory device
US5404327A (en) * 1988-06-30 1995-04-04 Texas Instruments Incorporated Memory device with end of cycle precharge utilizing write signal and data transition detectors
US5511029A (en) * 1993-05-25 1996-04-23 Mitsubishi Denki Kabushiki Kaisha Test circuit in clock synchronous semiconductor memory device
US5844849A (en) * 1995-05-22 1998-12-01 Mitsubishi Denki Kabushiki Kaisha Dynamic semiconductor memory device having fast operation mode and operating with low current consumption
US5959929A (en) * 1997-12-29 1999-09-28 Micron Technology, Inc. Method for writing to multiple banks of a memory device
US5978300A (en) * 1996-03-11 1999-11-02 Kabushiki Kaisha Toshiba Semiconductor memory having an overlaid bus structure
US6101139A (en) * 1998-02-19 2000-08-08 Micron Technology, Inc. Device and method for margin testing a semiconductor memory by applying a stressing voltage simultaneously to complementary and true digit lines
US6144598A (en) * 1999-07-06 2000-11-07 Micron Technology, Inc. Method and apparatus for efficiently testing rambus memory devices
US6285962B1 (en) * 1998-08-26 2001-09-04 Tanisys Technology, Inc. Method and system for testing rambus memory modules
US6339817B1 (en) * 1997-09-16 2002-01-15 Nec Corporation Semiconductor memory including main and sub memory portions having plural memory cell groups and a bidirectional data transfer circuit
US6377501B2 (en) * 1997-09-16 2002-04-23 Nec Corporation Semiconductor integrated circuit device
US6530045B1 (en) * 1999-12-03 2003-03-04 Micron Technology, Inc. Apparatus and method for testing rambus DRAMs

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5070297A (en) * 1990-06-04 1991-12-03 Texas Instruments Incorporated Full wafer integrated circuit testing device
US5675539A (en) * 1994-12-21 1997-10-07 Sgs-Thomson Microelectronics, S.A. Method and circuit for testing memories in integrated circuit form
US5629617A (en) * 1995-01-06 1997-05-13 Hewlett-Packard Company Multiplexing electronic test probe
JP3172760B2 (en) * 1997-03-07 2001-06-04 東京エレクトロン株式会社 Vacuum contactor
US6055199A (en) * 1998-10-21 2000-04-25 Mitsubishi Denki Kabushiki Kaisha Test circuit for a semiconductor memory device and method for burn-in test
JP3714044B2 (en) * 1999-07-15 2005-11-09 セイコーエプソン株式会社 LIQUID CRYSTAL DISPLAY DEVICE, ITS MANUFACTURING METHOD, AND ELECTRONIC DEVICE
US6988084B1 (en) * 2000-07-11 2006-01-17 F.E. Dibacco, Inc. Automated system for analyzing charges and credits of banks and other financial institutions

Patent Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5404327A (en) * 1988-06-30 1995-04-04 Texas Instruments Incorporated Memory device with end of cycle precharge utilizing write signal and data transition detectors
US5126973A (en) * 1990-02-14 1992-06-30 Texas Instruments Incorporated Redundancy scheme for eliminating defects in a memory device
US5511029A (en) * 1993-05-25 1996-04-23 Mitsubishi Denki Kabushiki Kaisha Test circuit in clock synchronous semiconductor memory device
US5844849A (en) * 1995-05-22 1998-12-01 Mitsubishi Denki Kabushiki Kaisha Dynamic semiconductor memory device having fast operation mode and operating with low current consumption
US5978300A (en) * 1996-03-11 1999-11-02 Kabushiki Kaisha Toshiba Semiconductor memory having an overlaid bus structure
US6339817B1 (en) * 1997-09-16 2002-01-15 Nec Corporation Semiconductor memory including main and sub memory portions having plural memory cell groups and a bidirectional data transfer circuit
US6377501B2 (en) * 1997-09-16 2002-04-23 Nec Corporation Semiconductor integrated circuit device
US5959929A (en) * 1997-12-29 1999-09-28 Micron Technology, Inc. Method for writing to multiple banks of a memory device
US6101139A (en) * 1998-02-19 2000-08-08 Micron Technology, Inc. Device and method for margin testing a semiconductor memory by applying a stressing voltage simultaneously to complementary and true digit lines
US6285962B1 (en) * 1998-08-26 2001-09-04 Tanisys Technology, Inc. Method and system for testing rambus memory modules
US6314036B1 (en) * 1999-07-06 2001-11-06 Micron Technology, Inc. Method and apparatus for efficiently testing RAMBUS memory devices
US6144598A (en) * 1999-07-06 2000-11-07 Micron Technology, Inc. Method and apparatus for efficiently testing rambus memory devices
US6530045B1 (en) * 1999-12-03 2003-03-04 Micron Technology, Inc. Apparatus and method for testing rambus DRAMs

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060085703A1 (en) * 2004-10-15 2006-04-20 Chang-Hyuk Lee Memory cell test circuit for use in semiconductor memory device and its method
US7363555B2 (en) 2004-10-15 2008-04-22 Hynix Semiconductor, Inc. Memory cell test circuit for use in semiconductor memory device and its method
US20090273991A1 (en) * 2008-04-30 2009-11-05 Seong-Hwi Song Semiconductor memory device, operating method thereof, and compression test method thereof
US8024627B2 (en) * 2008-04-30 2011-09-20 Hynix Semiconductor Inc. Semiconductor memory device, operating method thereof, and compression test method thereof

Also Published As

Publication number Publication date
US20040233738A1 (en) 2004-11-25
US20050262405A1 (en) 2005-11-24
US20070168790A1 (en) 2007-07-19
US7168018B2 (en) 2007-01-23
US6986084B2 (en) 2006-01-10
US6854079B1 (en) 2005-02-08
US6530045B1 (en) 2003-03-04

Similar Documents

Publication Publication Date Title
US7168018B2 (en) Apparatus and method for reducing test resources in testing DRAMs
US6550026B1 (en) High speed test system for a memory device
US5428575A (en) Semiconductor memory device with comparing circuit for facilitating test mode
US6684356B2 (en) Self-test ram using external synchronous clock
US6108252A (en) Integrated circuit memory devices having self-test circuits therein and method of testing same
US6389525B1 (en) Pattern generator for a packet-based memory tester
US6483760B2 (en) Semiconductor memory integrated circuit operating at different test modes
US6442724B1 (en) Failure capture apparatus and method for automatic test equipment
JP3736701B2 (en) Parallel test circuit for semiconductor memory device
US5533194A (en) Hardware-assisted high speed memory test apparatus and method
US7886206B2 (en) Semiconductor memory test device and method thereof
KR100487180B1 (en) Testing RAMBUS memories
US6484289B1 (en) Parallel data test for a semiconductor memory
US6158036A (en) Merged memory and logic (MML) integrated circuits including built-in test circuits and methods
US6740929B2 (en) Semiconductor device and method for testing semiconductor device
US5668764A (en) Testability apparatus and method for faster data access and silicon die size reduction
US20060136792A1 (en) Random access memory having test circuit
US6718487B1 (en) Method for high speed testing with low speed semiconductor test equipment
US8488400B2 (en) Multi-port memory device
US5986953A (en) Input/output circuits and methods for testing integrated circuit memory devices
US6701470B1 (en) Method for testing a memory device having different number of data pads than the tester
US6288957B1 (en) Semiconductor memory device having test mode and method for testing semiconductor therewith
KR100221073B1 (en) Synchronous semiconductor memory device
US6466495B2 (en) Electronic circuit, test-apparatus assembly, and method for outputting a data item
Wang A Universal Test Pattern Generator for DDR SDRAM

Legal Events

Date Code Title Description
FEPP Fee payment procedure

Free format text: PAYER NUMBER DE-ASSIGNED (ORIGINAL EVENT CODE: RMPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

FPAY Fee payment

Year of fee payment: 4

FEPP Fee payment procedure

Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

Free format text: PAYER NUMBER DE-ASSIGNED (ORIGINAL EVENT CODE: RMPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

REMI Maintenance fee reminder mailed
LAPS Lapse for failure to pay maintenance fees
STCH Information on status: patent discontinuation

Free format text: PATENT EXPIRED DUE TO NONPAYMENT OF MAINTENANCE FEES UNDER 37 CFR 1.362

FP Lapsed due to failure to pay maintenance fee

Effective date: 20150123